Canuck Audio Mart Hifi and Audio ForumCanuck Audio Mart Hifi and Audio Forum
It is currently Wed Jun 20, 2018 8:43 am

All times are UTC - 8 hours




Post new topic Reply to topic  [ 15 posts ] 
Author Message
PostPosted: Mon Feb 19, 2018 6:51 pm 
Offline

Joined: Thu Dec 14, 2017 3:22 pm
Posts: 32
Location: London, ON, CA
I have spent weeks learning about crossovers and trying various programs to design one. Here is what I came up with,

Image

Image

Does it look any good?

Thank you,

David.


Top
 Profile  
Reply with quote  
PostPosted: Mon Feb 19, 2018 9:10 pm 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
I'll bite but I'm no expert.
Congrats and welcome.
I haven't spent much time with XSim.
Did you measure the FRDs or copy/paste?
BSC and combined inductance looks about right for typical cabinet width.

Just a suggestion: I would combine the .5 and the 1.0 inductors to a 1.5. The "baffle block" doesn't affect the tweeter significantly.

The phase/summing seems off. Often on a flat baffle (I assume), the crossover needs to be asymmetric to match phase. (for example: 2nd order LP, 3rd order HP) and probably one out of polarity. Neither of those drivers have any requirement for such steep 4th order filters and I always try to start conservatively with as few components as possible.

Neat trick on the software is to invert polarity of the tweeter and see the reverse null. You can estimate the horizontal offsets of each driver by the depth to cone/dome meets the voice coil but that can also be measured and calculated.

The tweeter's rise and dip between 3K and 5K will probably be noticeable. Too high for the RS180s to raise the crossover point but you can use a series notch filter (cap/resistor/coil) on the HP. Should be workable if that's an 8 ohm tweeter.

Try exchanging the values for C1 and C2.

You'll get more assistance from Parts Express' TechTalk as long as you're humble over there.

edit: when I said horizontal offset, i should have said, z-offset.


Last edited by ngjockey on Wed Feb 21, 2018 12:50 am, edited 1 time in total.

Top
 Profile  
Reply with quote  
PostPosted: Tue Feb 20, 2018 9:06 pm 
Offline

Joined: Thu Dec 14, 2017 3:22 pm
Posts: 32
Location: London, ON, CA
Here is another stab at it, I changed the drivers to bring the cost down 60% in case this whole experiment is a bust.

I took the FRD and ZMA files from the Parts Express web site.

These will be up against a wall on a large flat surface, so I am not sure how much baffle compensation I will actually need.


Attachments:
da175 imp.jpg
da175 imp.jpg [ 141.2 KiB | Viewed 1172 times ]
da175 freq.jpg
da175 freq.jpg [ 147.09 KiB | Viewed 1172 times ]
da175 cross.jpg
da175 cross.jpg [ 105.05 KiB | Viewed 1172 times ]
Top
 Profile  
Reply with quote  
PostPosted: Wed Feb 21, 2018 12:31 am 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
I like the overall curve but woulda thought L2 seems small. Then again, those DA's have a bit of taper. Personally, I've gone away from Zobels as I rarely found them necessary. Never used a Zobel on a HP. Couple things strike me as odd there. First, the value of L1 seems high and that rolloff seems steep for 2nd order. Second, usually where the phase does that vertical overlap is closer to the crossover point. That's not a rule or anything, just what I'm used to seeing.

Looking at that DC28 really makes me miss the RS28's. Take a look at some of the Peerless, TB's or SB's for not that much more $. If the sensitivity is too high, that's an easy fix. Problem is they don't provide the FRD/ZMA files. There are programs like SPLTrace and others (FPGraphTracer?) that can help.

What I would do with that DC28 would be to use a higher crossover, 2 -2.5K, and a soft knee by adding a resistor in series with L1.

What happens when they are close to the back wall is a mid to upper bass suckout so you might need even more BSC. Even that doesn't help much because it's narrow.


Top
 Profile  
Reply with quote  
PostPosted: Wed Feb 21, 2018 4:38 am 
Offline

Joined: Thu Dec 14, 2017 3:22 pm
Posts: 32
Location: London, ON, CA
I ditched the Zobels.

The really are crossing over around 2K when you double the output of the woofer, the graph just doesn't look like it because the crossover point shows the output of just one woofer, not the sum of two.


Attachments:
da175 freq imp v2.jpg
da175 freq imp v2.jpg [ 142.56 KiB | Viewed 1079 times ]
da175 freq xo v2.jpg
da175 freq xo v2.jpg [ 145.64 KiB | Viewed 1079 times ]
da175 xo v2.jpg
da175 xo v2.jpg [ 109.36 KiB | Viewed 1079 times ]
Top
 Profile  
Reply with quote  
PostPosted: Wed Feb 21, 2018 7:04 pm 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
Ran it through WinPCD with 18mm z-offsets for the woofers. Been a couple years since I've done this and forgetting if that should be negative or positive offset. Positive, looks good if you add a 4.7 ohm shunt resistor in series after L4 and invert the tweeter polarity. Negative is about 90 degrees out of phase at the crossover, in which case, you would go 2nd/3rd.


Attachments:
LR.jpg
LR.jpg [ 7.78 KiB | Viewed 1023 times ]
Top
 Profile  
Reply with quote  
PostPosted: Wed Feb 21, 2018 9:55 pm 
Offline

Joined: Thu Dec 14, 2017 3:22 pm
Posts: 32
Location: London, ON, CA
The final crossover didn't change much.

I ordered everything tonight and will start building my first ever multi driver speaker with home made crossover!

If there is interest, I will post pics of the build as I go.

Thank you,

David.


Attachments:
da175 imp.jpg
da175 imp.jpg [ 142.67 KiB | Viewed 971 times ]
da175 curve.jpg
da175 curve.jpg [ 143.39 KiB | Viewed 971 times ]
da175 crossover done.jpg
da175 crossover done.jpg [ 108.69 KiB | Viewed 971 times ]
Top
 Profile  
Reply with quote  
PostPosted: Fri Feb 23, 2018 8:10 pm 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
Played around a bit more. Best I could do with that combo is a 1st order LP with a Zobel (6.8 mH, 10 uF/4.7 ohm) and a 1st order HP cap (5.6 uF) followed by a series padding resistor. Decent phase tracking. Not a lot of tweeter protection and not a lot woofer breakup suppression but I couldn't find a better compromise that didn't get weird. Naturally, impedance and reactance were smooth, as simple xovers tend to be.


Top
 Profile  
Reply with quote  
PostPosted: Fri Feb 23, 2018 8:29 pm 
Offline

Joined: Sat Sep 01, 2012 5:08 am
Posts: 628
Location: Montréal, QC, CA
ngjockey wrote:
Played around a bit more. Best I could do with that combo is a 1st order LP with a Zobel (6.8 mH, 10 uF/4.7 ohm) and a 1st order HP cap (5.6 uF) followed by a series padding resistor. Decent phase tracking. Not a lot of tweeter protection and not a lot woofer breakup suppression but I couldn't find a better compromise that didn't get weird. Naturally, impedance and reactance were smooth, as simple xovers tend to be.
Yes you are right to add the padding resistor because your low pass Zobel could create some 'ringing' on attacks. In my simulation, I can't do better (as 1st order) than you. Well done.

@DaveFred, interesting.
Post your impressions with your next pics. Thanks.

_________________
Nakamichi BX-300/Balance BW-1010 DAC, Balance MR-50R, Infinity RS 8 Kappa
Nakamichi Dragon/Onix Dac25b, Audio Research SP-9, Spectral DMA-100s/Nakamichi PA-7II, Martin-Logan CLSII/Balance Helicia


Top
 Profile  
Reply with quote  
PostPosted: Sat Mar 10, 2018 8:02 am 
Offline

Joined: Sun Sep 07, 2008 7:12 am
Posts: 91
Location: Regina, SK, CA
Do you have the ability to do frequency measurements? Using the data files from the manufacturer is very difficult to get an accurate simulation. At the least look at using Jeff Bagby's response blender to add baffle effects and generate minimum phase data. I think you will find it changes your computer model a great deal plus you can see the amount of BSC to correct for. Remember to add delay to the low pass sections for the acoustic offset.

D.


Top
 Profile  
Reply with quote  
PostPosted: Sun Mar 11, 2018 6:38 pm 
Offline

Joined: Sun Sep 07, 2008 7:12 am
Posts: 91
Location: Regina, SK, CA
This is my take on the xover after simulating on a 8 x 20.5" baffle.

Attachment:
dayton3.png
dayton3.png [ 234.2 KiB | Viewed 542 times ]


Attachment:
dayton1.png
dayton1.png [ 239.48 KiB | Viewed 542 times ]


Top
 Profile  
Reply with quote  
PostPosted: Sun Mar 11, 2018 8:19 pm 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
Deant, that works better. Might be a bit much BSC for the application, which I read (elsewhere) would be a nearfield desktop/computer setup. In which case, lowering the LP inductor down to 1.0 and raising the cap to 39 or 43 uF might be a compromise. Without in-box measurements, it's guesswork. Had some drivers that I needed to tame with the crossover equivalent of a sledgehammer.

BTW, reverse polarity on tweeter.


Top
 Profile  
Reply with quote  
PostPosted: Sun Mar 11, 2018 9:29 pm 
Offline

Joined: Sun Sep 07, 2008 7:12 am
Posts: 91
Location: Regina, SK, CA
Why reverse polarity on tweeter? In the model the drivers are no where close to 180deg out of phase with each other. I guessed at the relative acoustic offset and based on this set of conditions the polarity is correct.


Top
 Profile  
Reply with quote  
PostPosted: Sun Mar 11, 2018 10:08 pm 
Offline

Joined: Sat Dec 29, 2012 8:16 pm
Posts: 494
Location: New Westminster, BC, CA
Your design translated to WinPCD with +18mm z-offset on tweeter. No vertical offsets included. Nice reverse null.


Attachments:
d1.jpg
d1.jpg [ 177.14 KiB | Viewed 484 times ]
Top
 Profile  
Reply with quote  
PostPosted: Mon Mar 12, 2018 11:37 am 
Offline

Joined: Sun Sep 07, 2008 7:12 am
Posts: 91
Location: Regina, SK, CA
Ahhh, I see whats happening. When I did my computer model, the phase data was stripped off and recalculated during tail correction. Without taking actual measurements there is no way to know which data set is the correct one. Guess this is a prime example of the model can only be as good as the data its fed.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 15 posts ] 

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 0 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group